Investigating model biases in the cold sector of extratropical cyclones using ground-based and airborne observations

Dry intrusions (DIs) are slantwise descending airstreams in extratropical cyclones transporting high momentum from the upper troposphere towards the planetary boundary layer. Near the surface, this momentum transport can result in gale-force wind gusts that contribute to the damage caused by winter storms in Western Europe. The accurate representation of extreme near-surface winds remains a challenge: The underlying processes contributing to the momentum transport to the surface cover multiple scales and are therefore treated differently in models. Especially small-scale to micro-scale processes such as surface fluxes, the surface roughness variability and the sub-grid surface roughness are not explicitly resolved in current operational models. These processes require an accurate representation, as they affect the downward mixing of the cold and dry air of the DI into the atmospheric boundary layer.

 

In this study, we first present the dynamic and thermodynamic structure of the lower troposphere in the cold sector during DI events using radiosonde observations at three North Atlantic coastal stations. DI events are identified via backward trajectories based on ERA5 reanalysis data. Trajectories descending at least 400 hPa within 48 hours are classified as DI trajectories. Second, we document potential biases and flow-dependent errors in state-of-the-art operational numerical weather prediction models (ICOsahedral Nonhydrostatic (ICON), Model Integrated Forecasting System (IFS)). We compare operational radiosonde measurements with IFS and ICON model data by analyzing background and analysis departures, as well as analysis increments (difference between background and analysis) for vertical profiles of temperature, humidity and wind. Preliminary results suggest a dry bias in the lower troposphere and a temperature dipole structure within the planetary boundary layer during DI events in both models. The temperature dipole introduces a cold bias near the surface and a warm bias at the transition of the planetary boundary layer to the free troposphere. Based on targeted radiosonde observations of cold front passages from the recent NAWDIC campaign, we analyse the temporal evolution of these biases relative to the time of the cold front passage and investigate mechanisms leading to the model bias. For a detailed case study, the radiosonde observations are complemented by ground- and airborne wind lidar and water vapor measurements.